High levels of MEK5/ERK5 expression have been detected in various human tumors. However these observations did not rigorously establish a causal link between ERK5 activity and cancer. Therefore, to test the hypothesis that ERK5 acts as a tumor promoter, we developed a mouse model in which ERK5 could be temporally and specifically ablated in the epidermis of the skin. We discovered that ERK5-deficient mice were resistant to DMBA/TPA-induced tumorigenesis, thereby providing a genetic demonstration that signaling downstream of ERK5 is essential for ras-mediated tumor formation. More specifically, the loss of ERK5 in the epidermis prevented transcriptional activation of IL1 and caspase 1-mediated cleavage of pro-IL1β following TPA treatment of the skin. This precluded the recruitment of neutrophils in the dermis to support the hyperproliferation of keratinocytes expressing oncogenic Ras. The requirement of ERK5 in IL1β production was recapitulated in HaCat cells, a human immortalised keratinocyte cell line. This experiment confirmed that keratinocytes respond to TPA by secreting IL1β and provided evidence that this model is transferable to human tissue. In particular, the availability of a specific ERK5 inhibitor, XMD8-92, enabled us to test the significance of our results for cancer therapy. Like the deletion of the erk5 gene, continued treatment of the skin with XMD8-92, increased the efficacy doxorubin to decrease tumor growth induced by DMBA/TPA treatment. This implies that low dose chemotherapy can be used in conjunction with XMD8-92 to decrease tumor burden.